The D.C. Circuit Upheld EPA's PFAS Superfund Designation. 250,000 Property Deals a Year Need a Screen Nobody Sells.
On August 18, 2026, the D.C. Circuit upheld EPA's designation of PFOA and PFOS as CERCLA hazardous substances. Phase I reports are now expected to evaluate PFAS to preserve a buyer's liability defense. 250,000 commercial property deals a year need a standardized screen. Nobody sells one.
The Problem
On August 18, 2026, owning commercial property near the wrong industrial history got more dangerous. A federal appeals court upheld EPA's decision to put PFOA and PFOS, the best-known forever chemicals, under Superfund law. PFAS are per- and polyfluoroalkyl substances used since the 1940s in nonstick coatings, stain repellents, firefighting foam, and chrome plating. They do not break down, they accumulate in blood, and EPA links the two listed compounds to cancer and developmental harm, which a buyer could ignore during due diligence for decades. The law just caught up.
That ended in two moves. First, on April 19, 2024, EPA released the final rule designating PFOA and PFOS, including their salts and structural isomers, as hazardous substances under CERCLA (the Comprehensive Environmental Response, Compensation, and Liability Act, the federal Superfund law), the agency's first-ever use of its direct-listing authority under 42 U.S.C. § 9602(a) (published May 8, 2024, 89 FR 39124; effective July 8, 2024). Second, on August 18, 2026, a unanimous three-judge panel of the D.C. Circuit, Judges Millett, Katsas, and Rao, rejected every industry challenge in Chamber of Commerce v. EPA, No. 24-1193: EPA's statutory reading was correct, its cost-benefit notice was adequate, and its decision to regulate despite scientific uncertainty was not arbitrary. The petitions were denied and the designation is in force; en banc rehearing or Supreme Court review remain available within their deadlines, but buyers and lenders are operating under the rule today.
The legal consequences are the part that matters for this business, because under CERCLA, liability is strict, joint and several, and retroactive: a current owner can be on the hook for contamination deposited decades ago by someone else, even if the owner never touched the chemicals.
As HazMat Magazine's analysis of the designation puts it, the rule triggers mandatory release reporting, authorizes federal cleanup responses, and invokes that liability framework in full. One softening exists: courts allow apportionment where a defendant proves a reasonable basis for dividing the harm, the Burlington Northern divisibility defense, but joint and several is the starting presumption. The default is brutal.
The one escape hatch a buyer has is the landowner liability defense, which requires conducting "all appropriate inquiries" before purchase. AAI means a Phase I environmental site assessment that follows the ASTM E1527-21 standard (ASTM International's Phase I practice) and satisfies 40 CFR Part 312. The ASTM text did not change with the designation; what changed is the universe of CERCLA hazardous substances the standard must cover, so environmental practice groups now treat PFOA and PFOS evaluation as required for the inquiry to count. Leave PFAS out of the file and the defense has a hole no lender's counsel will ignore: no court has held a PFAS-only omission fatal to AAI yet, but the profession has moved. PFAS went from a polite non-scope consideration to a line item in the liability defense.
The scale of the transaction flow is well measured. EDR's industry surveys put the figure at about 250,000 commercial real estate deals a year affected by the Phase I standard, and most CRE transactions require a Phase I, per Partner Engineering and Science. A standard Phase I for a commercial property runs $2,000 to $5,000, averaging around $2,500, per AEI Consultants, which puts roughly $625 million a year through the Phase I workflow. The reports are, per EDR's own survey writeup, "often as varied and diverse as the consultants whose judgment is relied upon." Consultants scope PFAS engagement by engagement. LightBox's own research calls it a conundrum. Lenders get a PDF that says something defensible but not standardized, and the next lender in the chain can't compare it to anything: same report, different risk.
Meanwhile the liability pool behind those PDFs is enormous. The U.S. Chamber of Commerce, which sued EPA to block the designation and lost, commissioned a Monte Carlo model of PFOA and PFOS cleanup costs at existing non-federal Superfund sites: mean present value $17.4 billion at a 3% discount rate, with a 90% prediction interval of $10 billion to $27.2 billion. Treat that as an advocacy-commissioned upper bound, not a neutral forecast: it was built to argue the designation was too costly. The annualized private-party costs still run $700 million to $800 million a year, every year, for decades.
For scale, $17.4 billion is about $135 per American household.
Releases of a pound or more of PFOA or PFOS in a 24-hour period now trigger immediate reporting to the National Response Center. And the teeth are not only EPA's: LightBox's post-ruling analysis notes that states and private parties may now pursue qualifying cost-recovery or contribution claims under CERCLA. Sites that already received regulatory closure for other contaminants can face PFAS reopeners, with one caveat: EPA has said the designation will not by itself reopen National Priorities List (NPL)-deleted sites, so the live exposure is five-year-review add-ons and state-closure reopeners, which turn on settlement terms.
Why Now
The circuit-level overhang is gone. The industry's best argument against the designation died on August 18, 2026, when the D.C. Circuit denied every petition, which means buyers and lenders who were deferring PFAS diligence pending the outcome have no cover left at the circuit level. The 2024 rule is in force, with the court's full reasoning published; further appeals remain available within their deadlines, but the market is operating under the rule today. The waiting is over.
PFAS moved from non-scope to in-scope. Before July 8, 2024, a Phase I could treat PFAS as an emerging contaminant the buyer might optionally consider, the way asbestos and mold get handled; now PFOA and PFOS are CERCLA hazardous substances, and the AAI defense requires evaluating them. Every Phase I written before the designation is arguably stale on this point, which means refinancings and portfolio reviews are re-screening opportunities too.
The teeth are private, not just governmental. EPA's April 19, 2024 enforcement discretion memo says the agency will focus on major polluters and does not intend to pursue passive receivers like water systems, landfills, and farms. That memo does not bind private parties bringing cost-recovery claims, and it does not bind the states. As Dragun Corporation notes, the assurance "does not preclude third-party lawsuits or state-enforcement actions." Lenders know this. They are tightening requirements ahead of the regulators, not behind them. Fear is a sales cycle. This business monetizes liability anxiety, which deserves one honest sentence: the anxiety already exists, it is already priced into every deal, and a $228 standardized screen is the cheapest form it has ever taken. The design commitment behind the sentence: publish the flag rate and the false-positive rate every year, and let lenders audit both.
Reopeners multiply the market. Sites that received regulatory closure for other contaminants can be reopened for PFAS, which means the screen's customer is not only the buyer in a transaction. It is every current owner of a property near a former chrome plater, airport, or fire-training area who wants to know whether their closed file is about to reopen. That is a far larger addressable base than 250,000 annual deals.
The data is public and the tooling is finally cheap. EPA's PFAS Analytic Tools, state site inventories, and NAICS-coded business registries are all public, while geocoding, historical-map digitization, and report generation are commodity engineering, so the product is an integration and credibility play, not a research project, which is why a small team can build it in six months.
The Gap in the Market
| Player | What They Do | What's Missing |
|---|---|---|
| LightBox (formerly EDR) | The dominant environmental database vendor. Sells records-search reports to ESA firms through the LightBox Live workspace; has been steadily expanding its PFAS database coverage since 2018. | Sells data, not a determination. A LightBox report gives the consultant 400 pages of records; it does not produce a standardized PFAS risk categorization, a sampling decision, or a continuing-obligations record. They are the obvious acquirer, not the competitor. |
| ERIS | Environmental database reports for ESA shops, used by firms like A3 Environmental for their records searches. | Same as LightBox: raw records, no PFAS liability workflow, no defensible screen product. |
| ESA consulting firms (Partner ESI, AEI, Terracon, AECOM) | Perform the actual Phase I reports. Some, like AEI, have published PFAS scoping approaches. Partner ESI wrote the industry explainer on the ESA standard changes. | Hours-based and engagement-specific. AEI's own published method is case-by-case. A 40-person regional ESA firm cannot productize a screen across its staff, and its business model rewards billing the hours, not eliminating them. These firms are the channel: white-label the screen and let them mark it up. |
| Environmental law firms (Kirkland & Ellis, MGKF) | Publish client alerts on the designation's consequences and advise on liability structuring. | Advice, not software. They sell memoranda interpreting the rule at partner rates; they do not screen 250,000 properties. They are referral partners and expert reviewers of the screen's report language, not rivals. |
| Generic environmental data startups | A handful of proptech and climate-risk firms sell flood, fire, and generic contamination scores. | None owns the CERCLA-specific workflow: the AAI defense record, the PFAS source-class taxonomy, the sampling-decision guidance, and the post-close obligations tracker as one product. A flood score does not preserve a Superfund defense. |
The pattern: the data vendors sell records, the consultants bill hours, the law firms sell memos. The missing piece is a standardized, auditable PFAS screen that sits inside the Phase I workflow and follows the property after closing. The closest thing to the product is a consultant's judgment, which is exactly what a lender cannot audit. Audits need numbers.
The Solution
A PFAS liability screening platform with three modules that follow the property through the transaction and beyond:
1. The screen ($175 white-label / $295–$495 direct): The user enters a property address, and the platform pulls EPA's PFAS Analytic Tools data, state PFAS site inventories, federal facility and Department of Defense aqueous film-forming foam (AFFF) site lists, airport and fire-training-area records, landfill locations, and North American Industry Classification System (NAICS)-coded source classes (chrome platers, metal finishers, textile coaters, semiconductor fabs) within defined radii, then runs the property's historical uses, from fire insurance maps, city directories, and aerial imagery, through a PFAS lens: which prior uses are known PFAS handlers. Output is a low/moderate/high risk categorization with the evidence trail, a sampling-decision guide (screen more, or proceed), and a report formatted to drop into the Phase I as the PFAS appendix. Turnaround: minutes, not the two to four weeks a Phase I takes.
2. Lender portfolio monitoring ($5,000–$25,000/year): Banks with commercial real estate books need the same screen across thousands of properties, re-run when new data lands. A regional bank that financed 800 strip malls in 2019 has 800 un-screened PFAS exposures sitting on its balance sheet. The platform monitors the portfolio against updated source data and flags properties whose risk category changes. That is the product a chief risk officer buys, not a transaction team: different buyer, bigger check. Follow the money.
3. Continuing-obligations tracker ($49/property/month): Post-close, the platform maintains the record that preserves the buyer's CERCLA defense: land-use restriction compliance checklists, release-reporting reminders keyed to the 1-pound rule, and a documented log of reasonable steps, so that if a neighboring dry cleaner or plater ever becomes a Superfund site, the owner has a timestamped, third-party-maintained defense file. This is the module that converts a transaction product into an annuity and the dataset into an acquisition target.
The Price of the Defense: An Original Calculation
Start with the price of the defense: a buyer pays roughly $2,500 for a Phase I to preserve the innocent-landowner defense against a liability regime whose downside is measured in the billions. The PFAS component of that defense currently costs whatever the consultant feels like billing that week, produces a non-comparable writeup, and carries no ongoing obligations record. A standardized screen at a $228 blended price (70% white-label at $175, 30% direct at $295 to $495; $227.50 exact) would be under 10% of the Phase I price for the fastest-growing liability category in the report. Nobody prices a $2,500 insurance ritual and then quibbles over the $228 line item that covers the newest listed risk: cheap insurance, fully auditable.
Now the attach math. 250,000 Phase I reports a year times a 30% screen attach rate times a $227.50 blended price equals $17.1 million a year in transaction-screen revenue. That is the base business, and it is deliberately the boring part, and boring is good.
The interesting part is the funnel the screen controls. A screen that comes back moderate or high risk is, by design, a triage decision: this property needs a Phase II investigation, which involves actual soil and groundwater sampling and runs $5,000 to $15,000. Assume 12% of screens flag for follow-up and an $8,000 average Phase II: 75,000 screens × 12% × $8,000 = $72 million a year in downstream investigation spend the screen routes to sampling firms. The platform's cut is the referral or scheduling fee, $500 to $1,000 per job, or roughly $6.8 million a year captured at the midpoint, which makes the platform the toll booth.
The third leg is the one nobody is pricing at all. CERCLA's landowner defenses, innocent landowner, bona fide prospective purchaser, contiguous property owner, all require continuing obligations after closing: take reasonable steps to stop or prevent releases, comply with land-use restrictions, cooperate with response actions. Today that obligation is a paragraph in a PDF that nobody tracks. A post-close continuing-obligations tracker at $49 per property per month turns a one-time screen into a multi-year record that preserves the buyer's defense. If 20% of screened deals convert: 15,000 properties × $49 × 12 months = $8.8 million in ARR ($26.5 million in cumulative contract value over three years). That is the moat: the database of who is actively preserving their CERCLA defense, updated monthly, which no raw-data vendor has and no consultant can replicate without software: an annuity, not a project.
Revenue Model
| Revenue Stream | Amount | Notes |
|---|---|---|
| White-label screen (ESA firms) | $175/report | Firms mark up to $300+. 85%+ gross margin; COGS is compute plus public data, roughly $10–$30 per report. |
| Direct screen (buyers, brokers, lenders) | $295–$495/report | Tiered by property complexity. Same margin profile. |
| Lender portfolio monitoring | $5,000–$25,000/year | Per-bank subscription, priced by portfolio size. Annual, sticky. |
| Continuing-obligations tracker | $49/property/month | Post-close annuity. 20% screen-to-tracker conversion at scale. |
| Phase II referral / scheduling fee | $500–$1,000 per job | On the ~12% of screens that flag for sampling. The funnel monetization. |
| API | $0.50–$2/call | For proptech platforms and data vendors embedding the screen. Volume play. |
Unit economics at 5,000 screens a year: 3,500 white-label × $175 plus 1,500 direct × $350 = $1.14M ARR at 85%+ gross margin. Phase II referrals: 600 jobs × $750 = $450K/year. Tracker conversions: 1,000 properties × $49 × 12 = $588K ARR by end of year two. Three lender pilots at $15K = $45K. Year-2 blended revenue near $2.2M on a team of 8 to 10, with the tracker base compounding.
Market Size
TAM: 250,000 Phase I transactions a year × $227.50 blended screen = $57M/year in screen revenue, plus lender portfolio monitoring: 500 CRE-active regional banks × $10,000/year = $5M/year, for a combined software TAM of about $62M/year, before the roughly $6.8M in annual Phase II referral fees the screen routes and captures.
SAM: Beachhead on independent ESA firms and regional lenders in the PFAS-active states, Michigan, Minnesota, New Jersey, Massachusetts, California, Wisconsin, where state programs and litigation have already trained the market to care. Roughly 30% of national deal flow at the same 30% attach rate: 250,000 × 30% × 30% × $227.50 = $5.1M/year in screens.
SOM (year 3): 5,000 screens a year = $1.14M ARR on the screen line (3,500 white-label × $175 plus 1,500 direct × $350), about 22% of SAM's screen line, the only line measured the same way on both sides, plus referral and tracker revenue detailed above. Aggressive, but the wedge is the ESA firms themselves, 20 white-label partners whose existing deal flow is the distribution, plus direct sales to lenders who need the screen on every file.
Startup Costs
| Category | Cost | Notes |
|---|---|---|
| MVP: data pipeline, geocoding, report engine (6 months) | $180K | 2 engineers. EPA and state source ingestion, NAICS source-class mapping, PDF report generation. |
| Environmental counsel (report language review) | $40K | A Superfund lawyer reviews every report template so the screen's wording supports, rather than undermines, the AAI defense. |
| E&O insurance ($2M) | $30K/yr | Table stakes in the environmental consulting world. No ESA firm white-labels an uninsured vendor. |
| Data licensing and cleaning | $30K | Commercial historical-map and city-directory data where public sources have gaps. |
| ESA firm pilot program (20 firms) | $25K | Subsidized onboarding, report-format customization, testimonial capture in the six launch states. |
| Founder salaries (year 1) | $240K | 2 founders, modest. The CEO should be able to sell to environmental consultants; the CTO builds the pipeline. |
| Operating buffer | $40K | Hosting, compliance, customer support line. |
| Total | $585K |
Limitations
The $17.4 billion figure is advocacy math from the industry that sued to block the rule. The Chamber's Monte Carlo model was built to argue the designation was too costly, and its assumptions lean pessimistic. Treat it as an upper-bound illustration of the liability pool's scale, not a forecast. The honest range is wide: the study's own 90% interval runs $10 billion to $27.2 billion, and it covers only existing non-federal Superfund sites, not the far larger universe of ordinary commercial properties. Wide error bars. Honest ones.
EPA's enforcement discretion memo genuinely softens the urgency for the cleanest parcels. A buyer of a suburban office building with no industrial history, no nearby plater or airport, and clean state records faces a low-probability PFAS exposure, and EPA has said it will not be the agency's focus, which means the screen's attach rate on low-risk property types will run well below the 30% in the SAM math even though it still works as cheap insurance.
There are no enforceable federal cleanup standards for PFAS in soil. EPA has published advisory Regional Screening Levels, and as Bryan Cave's 2025 survey documents, states are filling the gap themselves with a patchwork of advisory, notification, and cleanup levels. Without a federal number, some buyers will treat a moderate-risk screen as a shrug: risk of what, exactly, at what concentration, enforced by whom. The screen can quantify proximity and history, but it cannot tell a buyer what a future cleanup will cost at their specific site. Proximity is not a price tag.
LightBox and ERIS could build this: they own the records-search relationships with every ESA firm in the country, and a PFAS screen is a natural upsell on their existing reports. The defense is that a data vendor's screen is a feature, while the continuing-obligations tracker and the lender portfolio product are a business. But if LightBox ships a credible screen bundled into its reports at $50, the white-label channel gets much harder. Speed to the ESA firms matters more than feature depth.
Then there is the conflict of interest at the center of the model, which a diligence product cannot leave unnamed. The platform decides which properties are moderate or high risk, and it earns $500 to $1,000 on roughly 12% of the screens it flags. That is a structural incentive to inflate the flag rate, and no list of mitigations dissolves it: the referral toll booth is the profit engine, and the conflict is structural and accepted as the price of the business model. What can be built around it is auditability, which is not absolution: price the screen as a fixed fee regardless of outcome, publish flag-rate benchmarks by property type so outliers are visible, disclose the referral economics inside the report itself, and keep the sampling arm at arm's length from the screening arm. A lender's counsel will ask about this in the first meeting; the only acceptable answer is the published numbers.
The false-positive problem is the mirror image of the conflict, and it falls on people who never asked for this product. A screen that cries moderate risk on a clean property can kill a sale, force a seller to fund a $5,000 to $15,000 Phase II, or blow up a refinancing. The platform's error rate is not an abstract quality metric; it is someone's collapsed deal. The honest countermeasures cost the business real money: publish the flag rate and the measured false-positive rate every year, give any seller who disputes a result a second independent screen at the platform's cost, and refund the screen fee plus the referral fee on any flag that independent sampling clears. A diligence product that cannot price its own mistakes is selling certainty it does not have.
The 250,000-deal figure reflects a functioning commercial real estate market. Deal volume is cyclical; in a downturn, Phase I counts fall and the attach base shrinks with them. The portfolio-monitoring and reopener products are the hedge, since they sell to owners and lenders regardless of transaction volume, but year-one revenue is still tied to deals closing.
Strongest Counterargument
Nobody will pay for this because PFAS diligence is already happening inside the Phase I, and a standalone screen is a solution in search of a buyer. Environmental consultants have been scoping PFAS for years, lenders already require Phase I reports that increasingly mention it, and EPA's own enforcement discretion memo signals the agency has no appetite for going after ordinary property owners. The August 2026 ruling changed the legal label, not the on-the-ground practice: good consultants were already evaluating PFAS as a business risk, the memo protects the sympathetic cases, and the marginal buyer of a clean office building will treat a $295 screen the way they treat the radon addendum, as an upsell to decline.
There is real force in this, and the honest version of the business concedes the first half. PFAS scoping is happening; it is just happening inconsistently, engagement by engagement, in a format no lender can compare across deals. LightBox's own researchers call it a conundrum handled case by case. That is precisely the condition that creates a standardization business: the work exists, the price is already being paid in consultant hours, and the output is not auditable. The screen does not invent a new cost; it productizes one buyers already incur, at a tenth of the price, in a format that survives a lender's file review. Standardization is the product.
On the enforcement memo, the counterargument misreads who the customer fears. EPA forbearance does not stop a neighboring property owner's cost-recovery suit, does not stop a state agency with its own PFAS program, and does not stop the lender who needs every file to document the same defensible standard. The teeth were never EPA chasing a strip-mall buyer. They are the private claim, the state action, and the refinancing that falls through because the Phase I's PFAS section was two sentences long. A standardized screen is what the lender's counsel points to when the file gets questioned. That is a product consultants cannot deliver by hand at scale.
What You Can Do
If you buy commercial property: Ask your environmental consultant, in writing, how the Phase I evaluates PFOA and PFOS under the current ASTM standard and whether the report documents the evidence for the conclusion, because a two-sentence PFAS section is a signal, not an analysis. If the property has any industrial history, or sits near a former chrome plater, airport, fire-training area, or landfill, budget for PFAS sampling in the Phase II scope rather than discovering it after closing.
If you lend on commercial property: Standardize the PFAS question across your portfolio before a borrower does it for you: pull the properties financed before July 2024 and check whether their Phase I reports evaluated PFOA and PFOS at all, since most did not, because they were written when PFAS was non-scope. The refinancing wave will surface this gap one file at a time unless you get ahead of it.
If you own or are selling commercial property: Get ahead of the buyer's screen. Commission your own PFAS screen before listing, so a moderate-risk flag arrives with your own sampling data attached rather than as a surprise inside the buyer's Phase I. Ask any screening vendor for its flag rate by property type and its measured false-positive rate; if it cannot produce both, its verdict is an opinion, not a measurement, and if a buyer's screen flags your property, insist on a second independent screen before anyone prices a Phase II into your sale.
If you run an ESA consulting firm: Productize the PFAS scoping you are already doing by hand. The firms that ship a consistent, branded PFAS appendix first will win the lender panels, because lenders buy comparability. Your competitors are not other consultants; they are the data vendors deciding whether to sell this as a $50 feature.
If you're building this: Start with 20 independent ESA firms in Michigan, New Jersey, and Minnesota, where state PFAS programs have already educated the market and the consultants feel the pain. White-label the screen and let the firms mark it up; they are your sales force and your credibility. Hire the Superfund lawyer before the second engineer: the report language is the product, and one badly worded determination that undermines a client's AAI defense ends the company. Build the continuing-obligations tracker in year one, not year three, because that is the dataset an acquirer pays for.
The Bottom Line
Two years ago PFAS was an emerging contaminant a buyer could politely skip; then EPA listed PFOA and PFOS under Superfund law, and on August 18, 2026, a federal appeals court told the industry's challengers that the listing stands. Overnight, a quarter-million commercial property deals a year gained a liability question that the existing machinery answers with engagement-priced hours and unauditable PDFs. CERCLA does not care that you never made the chemicals; it cares that you own the dirt. A $228 screen that turns the PFAS question into a standardized, defensible, auditable record, and then follows the property after closing to preserve the buyer's defense, is not a bet on more regulation. Paper beats vibes. It is a bet that 250,000 buyers a year would rather pay for the answer than inherit the question. The screen costs less than the Phase I's travel budget. That is a bet worth taking.